Submitted by mnannini on
Research and development is by nature uncertain. Meeting project milestones in university-industry projects can be a challenge. Any experiences or advices that can be shared on how to better manage R&D projects ?
Two ends of the spectrum
There are two ends of the spectrum of management techniques for R&D projects:
Biggest problem is that, even when you make very accurate or even relaxed average predictions of how long things take, nothing ever takes shorter than the time you put in the time schedule. Hence, the only options left are that it is done in time or too late. That is why R&D project are almost always over time.
If you have to go for option (2), it is good to read about LEAN project development and the Toyota product development system. They are counter-intuitive. And read "The Goal" too....
Let's not confuse R&D and Universities
The way to manage R&D projects is to treat them with the respect we treat our projects that are delivering value now: manage them tightly, and insist on time-lines, deadlines, and professional behaviors around effort and reporting.
R&D happens in all kinds of places. Universities are more difficult places to make things happen, in many ways, but let's not paint all R&D as being managed like it is at universities.
And, by the way, it's MUCH easier for R&D folks to fight the professionalization of their efforts...but it's still ineffective, and monumentally inefficient.
Any insight (practical application or research) on what are acceptable budget variances? For example, you are 25% the way through your R&D project, yet you have only spent 5% of your allocated budget. Is this acceptable?
Thanks for your insight.
Research frequently is non-linear in its spending. Commonly, research tasks are subcontracted or outsourced. Those payments can be a majority of the expenses. Outsources services are frequently billed in one or two payments. Track the largest items in the budget: staffing, external expenditures, and fixed overhead costs (bench fees, leases, or depreciation). Staffing and overhead tends to be linear and external spends can be non-linear. If large contracts have yet to be paid in the year, it is common to request that a portion of the payment be booked at the completion of a certain part of the work. This avoids the problem of having work contracted but the payment is only billed in Q4. Sometimes, budget cuts will target large open accounts like you describe. That can put the project in the red or push payment into the next year which then starts the project the hole.
If budget details are available for previous years, plotting year-to-date against earlier run rates should provide insight into the variances based on experience.
I am a project manager working in industrial R&D. Our projects go from concept to market, so have a lot of people involved with different drivers. The key thing I try to get across to the team is what a project is - it is shocking how few people that are not project managers do not get the fact that it is a beginning, a middle and an end. You should know roughly when all of these will happen before you start (the nearer they are the less rough). If, when you think you are ready to start a project, you do not have these facts as an absolute bear minimum then 1 of 2 things has happened: 1. You have not really conceived of the objective you need to achieve in the project, so you do not really understand the scope in enough detail; or 2. Actually you are embarking on a programme of work, and you have misunderstood the management requirements of that work, and not applied enough of a governance framework, this programme will almost certainly involve more than 1 project, all of which you should understand in principle, and have the nearest few to starting well defined. This does not have to be a major overhead, but it is important to recognize when you are doing a project, with deliverables and when you are doing a programme with outcomes.
The other thing that R&D projects/programmes need a lot of is planning. It is really easy to say well we can't plan in detail because we are going to find that out as we go - but certainly if it is a scientific project you should always enter a piece of work with a hypothesis as this forms the basis of your objective. The same should be true for engineering to a degree – you will have an expectation of success, which will inform the schedule of work and amount of repeat you might need to do.
When you are in a project there is no real difference to the management, but you will need more contingency, and if it is affects the P&L to keep you finance department well informed of successes and failures.
Manage them similarly to non r&d projects
I believe you can manage r&d projects similarly to how you manage other types of projects. Things also change in engineering projects and working plans and activities may need to be revised. The diference is it can happen due to external factors and not as a result of the work itself as it happens in research as one finds out the outcome as the work progresses. Perhaps in r&d projects the plan must be revised more often as with regardes to going in the directions it should be going, but it is still good to work with "who does what by when" and it is still good to have set milestones. Exactly whatever is achieved át. The milestones may depend on the result of the research but milestones should still be met. They may be a bit more general than in non-research projects i guess.